Technical Note. GeneChip 3 IVT PLUS Reagent Kit vs. GeneChip 3 IVT Express Reagent Kit Comparison. Introduction:

Transcription

1 Technical Note GeneChip 3 IVT PLUS Reagent Kit vs. GeneChip 3 IVT Express Reagent Kit Comparison Introduction: Affymetrix has launched a new 3 IVT PLUS Reagent Kit which creates hybridization ready target from total RNA samples for gene expression profiling analysis optimized for GeneChip 3' Expression Arrays. The kit is designed for sample types such as fresh or frozen tissues, cell lines and globin-reduced whole blood. It generates amplified and biotinylated complementary RNA (crna, also known as arna) from poly(a) RNA in a total RNA sample. RNA amplification is based upon linear amplification and employs T7 in vitro transcription (IVT) technology. The kit does not require any up-front removal of ribosomal RNA due to the oligo-d(t) priming strategy which primes the poly(a) tail, similar to the 3 IVT Express Kit. The 3 IVT PLUS Kit is comprised of reagents and a protocol for preparing hybridization-ready targets from 50 to 500 ng of total input RNA. Experimental Results: The experimental data presented herein is a comparison of the new 3 IVT PLUS Reagent Kit (P/N ) against the previous 3 IVT Express Reagent Kit (P/N ). We show performance equivalency for crna yields, technical reproducibility, signal correlation, and foldchange correlation (MAQC A and B samples only) using Affymetrix GeneChip Human Genome U133 Plus 2.0 (HG-U133 Plus 2.0) Arrays. Hybridization targets from both kits were prepared from the commercial total RNA samples listed in Table 1. Figure 1 shows the Agilent Bioanalyzer profiles and RNA Integrity Numbers (RIN) for each of the samples. The minimum input mass amount of 50 ng of total RNA and technical triplicates assays and arrays were performed. Following purification, crna yields were calculated based on concentration measurements from a Nanodrop spectrophotometer and crna size distributions were analyzed using an Agilent 2100 Bioanalyzer RNA 6000 Nano Kit. U133 Plus 2.0 arrays were hybridized with 50 ng/µl labeled targets per recommendations from both protocols. The resulting CEL files were quantile-normalized as a group in Expression Console 3.1 using the RMA algorithm for probeset summarization. The MAS 5.0 algorithm was used to calculate percent present (%P). Table 1: Total RNA samples used in this study RNA Description Vendor Part # MAQC A Universal Human Reference RNA Agilent MAQC B Human Brain Reference RNA Life Technologies AM6050 HeLa HeLa Cell Line Total RNA Life Technologies AM7852 LNCaP LNCaP Cell Line Total RNA Jinfiniti Biosciences LNCaP h. brain Human Brain Total RNA Life Technologies AM7962 h. liver Human Liver Total RNA Life Technologies AM7960 h. muscle Human Muscle Total RNA Life Technologies AM7982 h. placenta Human Placenta Total RNA Life Technologies AM7950

2 Average Total crna Yield (µg) Figure 1: Bioanalyzer images and RINs for the total RNA samples used in this study As seen in Figure 2, both kits generated similar crna yields from the 50 ng minimum input amount of total RNA which was sufficient for the 49- format array hybridization threshold for all samples. As seen in Table 2, the average technical reproducibility from the triplicates were greater than 0.97 and % Median Coefficient of Variation (% CV) were less than 10% which indicates excellent precision in the data. Also, % Present (% P) differed by less than 1% between kits which indicates similar detection sensitivity. Signal correlations plots for all probesets are shown in Figure 3. The signal correlations for all samples were greater than 0.95 which indicates excellent agreement between the two assays ' IVT PLUS 3' IVT Express MAQCA MAQCB HeLa LNCaP h. brain h. liver h. muscle h. placenta 50 ng Total RNA Input 16 hours IVT Incubation Figure 2: crna Yields

3 Table 2: Technical reproducibility, % Present, and % Median Coefficient of Variation Pairwise Replicate Signal Correlation (average ± standard deviation) 3' IVT 3' IVT PLUS Express %Present (average ± standard deviation) % Median CV Assay 3' IVT 3' IVT 3' IVT 3' IVT PLUS Express PLUS Express MAQC A ± ± ± ± % 7.43% MAQC B ± ± ± ± % 7.99% HeLa ± ± ± ± % 5.95% LNCaP ± ± ± ± % 7.37% h. brain ± ± ± ± % 7.60% h. liver ± ± ± ± % 8.51% h. muscle ± ± ± ± % 8.36% h. placenta ± ± ± ± % 7.02% Figure 3: Signal Intensity Correlation of All Probesets Each scatter plot shows the average RMA summarized probeset signal intensities (log 2 ) comparing the 50 ng input of the 3 IVT PLUS Kit (y-axis) to the 3 IVT Express Kit (x-axis). Since MAQC A and B samples represent a wide range of both signal and fold-change differences, we show a more in depth look at the data in the analyses that follow. As seen in Figure 4, crna lengths from the new 3 IVT PLUS Kit were slightly longer than the 3 IVT Express Kit. This gives one confidence in the data integrity since 3 IVT arrays have probesets which are, by design, closer to the 3 -end of transcripts. Summary of the metrics derived from Expression Console analysis is shown in Table 3. Overall, %P and CV metrics were comparable between the two assays while the Actin 3 /5 bias ratio is higher for 3 IVT Express Kit, in agreement with the Bioanalyzer profiles.

4 Figure 5 shows the scatter plot of average fold-change values [log(maqc A/B)] comparing the 50 ng input of the 3 IVT PLUS Kit (y-axis) to the 3 IVT Express Kit (x-axis). The Pearson Correlation Coefficient (r) calculated without filtering is and the fold-change discordance (significant p- value < and fold change > 1) is 0.44%. This demonstrates that not only were the absolute signal intensities from both assays in good agreement (Figure 3) but the ability to detect meaningful gene-level differences between two samples remains very similar. Figure 4: Assay-derived crna profiles from MAQC A and B samples Average size nucleotide [nt] was determined from 50% of the total area with region start size at 60 nt. Table 3: Expression Console analysis summary Kit 3 IVT PLUS 3 IVT Express RNA Input 50ng MAQCA 50ng MAQCB 50ng MAQCA 50ng MAQCB average ± stdev average ± stdev average ± stdev average ± stdev % Median CV 6.43% 7.29% 7.43% 7.99% % Present (MAS5) ± ± ± ± 0.50 Actin 3 /5 Ratio 1.30 ± ± ± ± 0.70 GAPDH 3 /5 Ratio 0.97 ± ± ± ± 0.08 Poly-A 3 Detection all P all P all P all P Poly-A Ratio r 2 * *3 signal intensities vs. relative ratio for each of the Poly-A Controls

5 Pearson r = Figure 5: MAQC A/B Fold-Change Correlation Conclusion: The 3 IVT PLUS and 3 IVT Express comparison data shows very high concordance on all the performance metrics for the two kits. The protocol and the workflow aspects for both the kits are extremely similar which will allow easy implementation of the new assay. This should enable users to seamlessly convert to the 3 IVT PLUS assay for their 3 Expression analysis target preparation needs.